A numerical rainfall forecast analysis method, device and medium

Abstract

The present application discloses a numerical rainfall forecast analysis method, device and medium, which are used to solve the problems that the existing technology cannot intuitively describe rainfall data, the national rainfall data is slow to load, and the rainfall numerical forecast deviation is difficult to estimate. The method includes: collecting boundary data of the area of ​​interest through multiple data sources, and converting the data format of the boundary data into a specified format; clipping the boundary data in the specified format to determine the data of interest in the area of ​​interest, and generating a corresponding data set; based on the data of interest in the data set, determining the rainfall prediction data corresponding to the time period to be executed, and obtaining the actual rainfall data corresponding to the time period to be executed, so as to determine the deviation data between the rainfall prediction data and the corresponding actual rainfall data for each latitude and longitude point; based on the deviation data, generating the root mean square difference data of the entire time period, so as to realize the numerical rainfall forecast analysis based on the root mean square difference data.

Description

Technical Field

[0001] The present application relates to the technical field of rainfall forecast analysis, and in particular to a numerical rainfall forecast analysis method, device and medium. Background Art

[0002] At present, with the continuous development of meteorological technology, the accuracy and timeliness of rainfall forecasts are particularly important in many fields such as agriculture, aviation, and urban planning. As a widely used data format, Network Common Data Form (NetCDF) plays an important role in the fields of meteorological, oceanographic, and climate data. However, radar echo data in NetCDF format is not intuitive when describing rainfall information, and it is difficult for ordinary users to interpret it.

[0003] Traditional NetCDF data processing methods usually involve complex data parsing processes, and these data are not easy to understand for ordinary users. In addition, due to the huge amount of rainfall data across the country, directly loading and processing these data will lead to inefficiency and slow loading speed, which is not conducive to real-time or quasi-real-time rainfall forecast services. At the same time, the accuracy of rainfall forecasts is a key indicator for evaluating the quality of a meteorological service. However, existing systems often lack effective analysis of forecast deviations, making it difficult for users to evaluate the reliability of forecast results. This uncertainty not only affects users' trust in rainfall forecasts, but also limits the application of meteorological services in certain high-precision scenarios. Summary of the invention

[0004] The embodiments of the present application provide a numerical rainfall forecast analysis method, device and medium to solve the technical problems in the prior art that rainfall data cannot be described intuitively, national rainfall data is slow to load, and rainfall numerical forecast deviation is difficult to estimate.

[0005] On the one hand, an embodiment of the present application provides a numerical rainfall forecast analysis method, comprising:

[0006] Collect boundary data of the area of ​​interest through multiple data sources, and convert the data format of the boundary data into a specified format;

[0007] Clipping the boundary data in a specified format to determine the data of interest within the region of interest, and generating a corresponding data set;

[0008] Based on the interest data in the data set, determine the rainfall prediction data corresponding to the time period to be executed, and obtain the actual rainfall data corresponding to the time period to be executed, so as to determine the deviation data between the rainfall prediction data and the corresponding actual rainfall data for each longitude and latitude point;

[0009] Based on the deviation data, root mean square difference data for the entire time period is generated, so as to realize numerical rainfall forecast analysis based on the root mean square difference data.

[0010] In one implementation of the present application, collecting boundary data of the region of interest through multiple data sources specifically includes:

[0011] When a preset scheduled task is triggered, the original NetCDF data is obtained based on web crawler technology, API interface call or geographic information service provider, and the original NetCDF data is stored in a pre-deployed OSS component;

[0012] The original NetCDF data is obtained from the OSS component to decompress the original NetCDF data to generate a corresponding sparse matrix, and based on the sparse matrix, boundary data corresponding to the region of interest is read.

[0013] In one implementation of the present application, converting the data format of the boundary data into a specified format specifically includes:

[0014] Cleaning the boundary data to remove duplicate data, invalid data or erroneous data in the boundary data, and determining a data format corresponding to the cleaned boundary data;

[0015] The data format of the cleaned boundary data is converted into GeoJson format and wkt format, and the cleaned and format-converted boundary data is stored in a database.

[0016] In an implementation of the present application, the clipping of the boundary data in the specified format to determine the interest data in the region of interest and generating a corresponding data set specifically includes:

[0017] Based on the boundary data in the specified format, a cropping operation is performed on the region of interest to extract the data of interest in the region of interest;

[0018] The intercepted interest data are sorted to generate a data set corresponding to the area of ​​interest; wherein the data set includes rainfall-related data corresponding to each latitude and longitude point in the area of ​​interest;

[0019] The rainfall-related data are verified, and according to the verification result, the rainfall-related data are corrected to obtain a standard data set.

[0020] In one implementation of the present application, determining the rainfall forecast data corresponding to the time period to be executed based on the interest data in the data set, and obtaining the actual rainfall data corresponding to the time period to be executed, specifically includes:

[0021] Receiving a time period selection instruction input by a user through a preset visual interface, and determining a time period to be executed corresponding to the time period selection instruction; wherein the length of the time period to be executed does not exceed two hours;

[0022] Based on the interest data in the standard data set, obtain the rainfall forecast data within the time period to be executed from the corresponding data source, and obtain the actual rainfall data corresponding to the time period to be executed from the actual observation site or the public rainfall data service;

[0023] The rainfall prediction data and the actual rainfall data are temporally and spatially matched to determine the corresponding relationship between the rainfall prediction data and the actual rainfall data.

[0024] In one implementation of the present application, performing spatiotemporal matching on the rainfall prediction data and the actual rainfall data to determine the corresponding relationship between the rainfall prediction data and the actual rainfall data specifically includes:

[0025] The rainfall prediction data and the actual rainfall data are divided according to the same time granularity to determine the rainfall prediction data and the actual rainfall data corresponding to each time point;

[0026] The rainfall prediction data and the actual rainfall data that are matched in time are spatially interpolated to determine the rainfall prediction data and the actual rainfall data corresponding to each spatial position, and the corresponding relationship between the rainfall prediction data and the actual rainfall data is determined in the case of time matching and space matching.

[0027] In one implementation of the present application, determining the deviation data between the rainfall prediction data and the corresponding rainfall actual data for each latitude and longitude point specifically includes:

[0028] Based on the correspondence between the rainfall prediction data and the actual rainfall data, the rainfall prediction data and the actual rainfall data at each latitude and longitude point at the same time point are determined, and the deviation data between the rainfall prediction data and the actual rainfall data matched in time and space are calculated; wherein the deviation data includes a difference index and a ratio index;

[0029] Comparing the deviation data with a preset deviation threshold to determine a corresponding magnitude relationship, and determining whether the deviation data is abnormal based on the magnitude relationship;

[0030] If so, the outliers in the deviation data are removed, and the deviation data after the outliers are removed is smoothed to obtain the target deviation data.

[0031] In one implementation of the present application, generating root mean square error data for the entire time period based on the deviation data, so as to implement numerical rainfall forecast analysis based on the root mean square error data, specifically includes:

[0032] Based on the deviation data, the root mean square difference data of the entire time period is calculated, and the root mean square difference data of the entire time period is counted, so as to visualize the root mean square difference data in a preset manner; wherein the preset manner includes a line chart and a bar chart;

[0033] Determining a change trend of the root mean square difference data, so as to analyze the numerical rainfall forecast based on the change trend;

[0034] According to the analysis results, the numerical rainfall forecast is optimized, and a corresponding analysis optimization report is generated, so as to send the analysis optimization report to the management personnel.

[0035] On the other hand, the embodiment of the present application further provides a numerical rainfall forecast analysis device, the device comprising:

[0036] at least one processor;

[0037] and, a memory communicatively coupled to the at least one processor;

[0038] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a numerical rainfall forecast analysis method as described above.

[0039] On the other hand, an embodiment of the present application further provides a non-volatile computer storage medium storing computer executable instructions, which, when executed, implement a numerical rainfall forecast analysis method as described above.

[0040] The present application provides a numerical rainfall forecast analysis method, device and medium, which at least have the following beneficial effects:

[0041] By collecting boundary data of the area of ​​interest from multiple data sources, it is possible to obtain regional information more comprehensively, reduce data omissions, and improve the efficiency and accuracy of data collection; converting the data format of boundary data into a specified format is conducive to subsequent data processing and analysis, ensuring that data from different sources can be seamlessly connected, and improving data compatibility and availability; clipping boundary data in a specified format can accurately extract the data of interest in the area of ​​interest, eliminate the interference of irrelevant information, and make subsequent analysis more focused on key areas; generating corresponding data sets provides rich data support for rainfall prediction and other related analyses, which helps to improve the depth and breadth of analysis; determining rainfall prediction data based on the data of interest in the data set and comparing it with the actual rainfall data can timely discover the deficiencies of the rainfall prediction model and provide data support for improving the accuracy of rainfall prediction; by calculating the deviation data between the rainfall prediction data of each longitude and latitude point and the corresponding actual rainfall data, it is possible to accurately identify in which areas the rainfall prediction model has deviations; generating root mean square deviation data for the entire time period realizes the numerical evaluation of rainfall forecast performance, making the forecast effect more intuitive and quantifiable. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0043] Figure 1 A flowchart of a numerical rainfall forecast analysis method provided in an embodiment of the present application;

[0044] Figure 2 A schematic diagram of the internal structure of a numerical rainfall forecast analysis device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0046] The embodiment of the present application provides a numerical rainfall forecast analysis method, device and medium, which can collect boundary data of an area of ​​interest through multiple data sources, so as to obtain regional information more comprehensively, reduce data omissions, and improve the efficiency and accuracy of data collection; convert the data format of the boundary data into a specified format, which is conducive to subsequent data processing and analysis, ensure that data from different sources can be seamlessly connected, and improve the compatibility and availability of data; crop the boundary data in the specified format, so as to accurately extract the data of interest in the area of ​​interest, eliminate the interference of irrelevant information, and make the subsequent analysis more focused on the key area; generate corresponding data sets, provide rich data support for rainfall prediction and other related analyses, and help to improve the depth and breadth of analysis; determine rainfall prediction data based on the data of interest in the data set, and compare it with the actual rainfall data, so as to timely discover the deficiencies of the rainfall prediction model and provide data support for improving the accuracy of rainfall prediction; by calculating the deviation data between the rainfall prediction data of each longitude and latitude point and the corresponding actual rainfall data, it is possible to accurately identify in which areas the rainfall prediction model has deviations; generate the root mean square difference data of the whole time period, realize the numerical evaluation of rainfall forecast performance, and make the forecast effect more intuitive and quantifiable. The present invention solves the technical problems that the existing technology cannot intuitively describe rainfall data, the loading of national rainfall data is slow, and the deviation of rainfall numerical forecast is difficult to estimate.

[0047] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0048] Figure 1 A flowchart of a numerical rainfall forecast analysis method provided in an embodiment of the present application.

[0049] The analysis method involved in the embodiments of the present application can be implemented by a terminal device or a server, and the present application does not impose any special restrictions on this. For the convenience of understanding and description, the following embodiments are described in detail by taking a server as an example.

[0050] It should be noted that the server may be a single device or a system consisting of multiple devices, that is, a distributed server, and this application does not make any specific limitation on this.

[0051] like Figure 1 As shown, a numerical rainfall forecast analysis method provided in an embodiment of the present application includes:

[0052] 101. Collect boundary data of the area of ​​interest through multiple data sources, and convert the data format of the boundary data into a specified format.

[0053] Specifically, when the server detects the triggering of a preset scheduled task, it obtains the original NetCDF data based on web crawler technology, API interface call or geographic information service provider, and stores the original NetCDF data in a pre-deployed OSS component; it obtains the original NetCDF data from the OSS component to decompress the original NetCDF data, generate a corresponding sparse matrix, and reads the boundary data corresponding to the area of ​​interest based on the sparse matrix.

[0054] In one embodiment, the system sets a daily scheduled task to automatically obtain meteorological data. When the system detects that the scheduled task is triggered, it will automatically start a web crawler to crawl data from a specific meteorological data website. The crawler locates and downloads the latest NetCDF format meteorological data file on the website according to preset rules. After the download is complete, the system will automatically store these original NetCDF data in the pre-deployed OSS (object storage service) component for subsequent processing and analysis.

[0055] The system then retrieves these raw NetCDF data files from the OSS component and decompresses them using a specialized decompression tool. After decompression, the system converts the data into a sparse matrix data structure that can efficiently store and process large amounts of meteorological data. Based on this sparse matrix, the system can quickly read the boundary data of a specific area of ​​interest to the user, such as a city or region, in preparation for subsequent data analysis and visualization.

[0056] In one embodiment, the system has established a cooperative relationship with a professional meteorological data service provider, and obtains meteorological data in NetCDF format through the API interface provided by the service provider. The system also sets a scheduled task. When the task is triggered, the system automatically calls the API interface and passes the necessary parameters to obtain the required data, such as time range, data type, etc.

[0057] After the API interface responds, the system receives the original NetCDF data file and stores it in the pre-configured OSS component. Similar to Example 1, the system then retrieves these data files from OSS, decompresses and converts them, and generates a sparse matrix to represent these meteorological data. Through this sparse matrix, the system can efficiently extract the boundary data of specific areas of concern to users, providing an accurate data basis for subsequent meteorological analysis and prediction.

[0058] In one embodiment of the present application, the server cleans the boundary data to remove duplicate data, invalid data or erroneous data in the boundary data, and determines the data format corresponding to the cleaned boundary data; converts the data format of the cleaned boundary data into GeoJson format and wkt format, and stores the cleaned and format-converted boundary data in a database.

[0059] In one embodiment, in a smart city project, it is necessary to integrate the boundary data of various areas of the city from multiple data sources. These data are crucial for urban planning, resource allocation, and emergency response. First, the boundary data of the city is obtained from different data sources. These data may contain duplicate, invalid, or erroneous information, so they need to be cleaned. Using a data deduplication algorithm, the unique identifier or key attributes of each boundary data, such as the latitude and longitude coordinate sequence, are compared to identify and remove identical duplicate records. For incomplete or obviously illogical data, for example, coordinate points that exceed the actual geographic range, data fields are missing, etc., they are marked as invalid and removed from the data set. For identifiable erroneous data, such as obvious coordinate point entry errors, we correct them according to other reliable data sources or preset rules. If correction is not possible, it is treated as invalid data.

[0060] After cleaning, a clean and accurate boundary data set is obtained. Next, this data needs to be converted into a specific format for subsequent spatial analysis and visualization. The cleaned boundary data is first checked and its original data format is determined, which may be Shapefile, KML or any other spatial data format. In order to meet the needs of various application scenarios, the cleaned boundary data is converted into two widely used spatial data formats-GeoJson and WKT (Well-Known Text). The GeoJson format is suitable for Web map services and lightweight data exchange, while the WKT format is easy to process and analyze in GIS software. Finally, a database system that supports spatial data query and indexing, such as PostGIS, is selected, so that these cleaned and converted boundary data are stored in the database to ensure efficient data retrieval and analysis performance.

[0061] 102. Crop the boundary data in the specified format to determine the data of interest in the region of interest, and generate a corresponding data set.

[0062] Specifically, in one embodiment of the present application, the server performs a cropping operation on the region of interest based on the boundary data in a specified format to extract the data of interest in the region of interest; and organizes the extracted data of interest to generate a data set corresponding to the region of interest. It should be noted that the data set in the embodiment of the present application includes rainfall-related data corresponding to each latitude and longitude point in the region of interest.

[0063] The server verifies the rainfall-related data and, based on the verification result, modifies the rainfall-related data to obtain a standard data set.

[0064] In one embodiment, it is assumed that the boundary data of the region of interest (such as a specific city or county) in GeoJson or WKT format has been obtained from the database. Now, it is necessary to cut out the rainfall data of the region of interest from a map containing national rainfall data.

[0065] First, use GIS software or corresponding programming libraries to load the boundary data of the area of ​​interest and the national rainfall data, such as Geopandas, Shapely, etc. The rainfall data may be stored in NetCDF, GeoTiff or other spatial data formats. Then, use the boundary data to clip the national rainfall data. This usually involves spatial overlay analysis, that is, only retaining rainfall data points that are within the boundaries of the area of ​​interest.

[0066] After clipping, a subset containing only rainfall data in the area of ​​interest is obtained. From this subset, the rainfall data for each latitude and longitude point is extracted and sorted. This may include resampling, interpolation, or aggregation of the data to ensure data consistency and availability. The sorted data is organized into a structured dataset, where each row represents a latitude and longitude point in the area of ​​interest, containing the latitude and longitude coordinates of the point and the corresponding rainfall data, such as rainfall amount, rainfall duration, etc.

[0067] After the data set is generated, it needs to be quality checked and corrected as necessary. By comparing the data of adjacent points, checking the spatiotemporal continuity of the data, or applying other statistical methods, possible outliers or erroneous data can be found and marked. For erroneous or abnormal data found in the verification, corrections are made based on the data of surrounding points, historical data, or other reliable information sources. For example, for obvious outliers, interpolation methods can be used or replacements can be made based on historical averages. After verification and correction, a higher quality and more reliable standard data set can be obtained for subsequent analysis and application.

[0068] 103. Based on the interest data in the data set, determine the rainfall prediction data corresponding to the time period to be executed, and obtain the actual rainfall data corresponding to the time period to be executed, so as to determine the deviation data between the rainfall prediction data and the corresponding actual rainfall data for each longitude and latitude point.

[0069] Specifically, the server receives the time period selection instruction input by the user through a preset visual interface, and determines the time period to be executed corresponding to the time period selection instruction. It should be noted that the length of the time period to be executed in the embodiment of the present application does not exceed two hours.

[0070] Based on the interest data in the standard dataset, the server obtains the rainfall forecast data for the time period to be executed from the corresponding data source, and obtains the actual rainfall data corresponding to the time period to be executed from the actual observation site or the public rainfall data service; the rainfall forecast data and the actual rainfall data are matched in time and space to determine the corresponding relationship between the rainfall forecast data and the actual rainfall data.

[0071] In one embodiment, first, a user-friendly visual interface is designed that allows the user to select a specific time period to query rainfall data. This interface can be part of a web application, mobile application, or desktop application. The user enters the time period of interest through a time selector on the interface, such as a date and time slider. The system verifies the user's selection to ensure that the selected time period does not exceed two hours, which can ensure data accuracy and processing efficiency. The system determines the time period to be executed based on the user's selection and uses it as the basis for subsequent data retrieval and processing.

[0072] The system first accesses the previously generated and stored standard datasets to extract the data of interest related to the time period and area of ​​interest selected by the user. According to the time period selected by the user, the system obtains rainfall forecast data from the corresponding data source, such as meteorological models, weather forecast APIs, etc. These data usually include information such as predicted rainfall amount, rainfall time and location. At the same time, the system obtains actual rainfall data for the same time period from actual observation sites or public rainfall data services, such as observation data from the China Meteorological Administration, satellite rainfall observations, etc.

[0073] Before matching, the system performs necessary preprocessing on the acquired rainfall forecast data and actual data, including data format conversion, coordinate system unification, time stamp alignment, etc., to ensure comparability between the two. The system uses a spatiotemporal matching algorithm to determine the correspondence between rainfall forecast data and actual data, mainly involving matching based on geographic location and timestamps, as well as considering factors such as the duration and intensity of rainfall events. After the matching is completed, the system outputs a report or visual chart containing the matching results, showing the comparison between the predicted rainfall and the actual rainfall, which helps users evaluate the accuracy of rainfall forecasts and provides support for subsequent meteorological analysis or decision-making.

[0074] In one embodiment of the present application, the server divides the rainfall prediction data and the actual rainfall data according to the same time granularity to determine the rainfall prediction data and the actual rainfall data corresponding to each time point; performs spatial interpolation on the time-matched rainfall prediction data and the actual rainfall data to determine the rainfall prediction data and the actual rainfall data corresponding to each spatial position, and determines the correspondence between the rainfall prediction data and the actual rainfall data when both time and space are matched.

[0075] In one embodiment, first, a suitable time granularity is determined, such as 5 minutes, 10 minutes or half an hour, etc., depending on the accuracy of the data and the analysis requirements. According to the selected time granularity, the rainfall prediction data and the actual rainfall data are divided, which means that the continuous data stream is divided into multiple time periods, each time period corresponding to a specific time point. At each time point, the rainfall prediction data and the actual rainfall data are extracted respectively. In this way, it can be ensured that the two are completely matched in time.

[0076] In order to match spatially, a spatial grid is first defined that covers the area of ​​interest. The size of the grid can be set according to the resolution of the data and the accuracy requirements of the analysis. For each spatial grid point, a spatial interpolation method is used to estimate the rainfall forecast data and actual rainfall data at that point, such as inverse distance weighting, Kriging interpolation, etc. In this way, even if the spatial resolution of the observation site or the output of the prediction model is different, the data corresponding to each spatial location can be obtained. Through spatial interpolation, it can be ensured that there are corresponding rainfall forecast data and actual rainfall data at each spatial grid point, thereby achieving spatial matching.

[0077] When both time and space match, the rainfall forecast data and actual rainfall data at the same time point and spatial location can be directly compared. By calculating the difference between the forecast value and the actual value, the accuracy of the rainfall forecast can be evaluated, such as absolute error, relative error, etc. Finally, a complete correspondence table or database is established, which records the rainfall forecast data and actual rainfall data at each time point and each spatial location, as well as the difference between them.

[0078] In one embodiment of the present application, the server determines the rainfall prediction data and the actual rainfall data at the same time point for each latitude and longitude point based on the correspondence between the rainfall prediction data and the actual rainfall data, and calculates the deviation data between the rainfall prediction data and the actual rainfall data that are matched in time and space. It should be noted that the deviation data in the embodiment of the present application includes a difference index and a ratio index.

[0079] The server compares the deviation data with the preset deviation threshold to determine the corresponding size relationship, and determines whether the deviation data has an abnormality based on the size relationship; if so, the server removes the abnormal values ​​in the deviation data and smoothes the deviation data after the abnormality is removed to obtain the target deviation data.

[0080] In one embodiment, first, according to the previously determined correspondence between the rainfall prediction data and the actual rainfall data, for each latitude and longitude point and the same time point, the corresponding rainfall prediction data and actual data are extracted. For each latitude and longitude point and time point, the deviation between the rainfall prediction data and the actual data is calculated. This includes calculating the difference and ratio between the two, i.e., the difference index and the ratio index. The difference index is the absolute difference between the predicted value and the actual value, while the ratio index is the ratio of the predicted value to the actual value.

[0081] Set preset deviation thresholds based on historical data and business needs. These thresholds can be fixed or adjusted based on dynamic changes in data. Compare the calculated deviation data with the preset deviation thresholds, including comparing whether the difference index and ratio index exceed the set threshold range. If the deviation data of a certain latitude and longitude point and time point exceeds the preset threshold, then these data points are considered outliers.

[0082] Once outliers are detected, these abnormal data points are removed from the deviation data set. After removing the outliers, the remaining deviation data is smoothed. This can be achieved by applying moving average, exponential smoothing or other statistical techniques to reduce the volatility of the data and highlight long-term trends. After the above processing, a set of cleaned and smoothed target deviation data is obtained, which more accurately reflects the difference between rainfall forecast and reality, and removes the interference of outliers.

[0083] 104. Generate root mean square error data for the entire time period based on the deviation data, and realize numerical rainfall forecast analysis based on the root mean square error data.

[0084] Specifically, in one embodiment of the present application, the server calculates the root mean square difference data of the entire time period based on the deviation data, and performs statistics on the root mean square difference data of the entire time period, so as to visualize the root mean square difference data in a preset manner. It should be noted that the preset manner in the embodiment of the present application includes a line chart and a bar chart.

[0085] The server determines the changing trend of the RMS error data, and analyzes the numerical rainfall forecast based on the changing trend; optimizes the numerical rainfall forecast according to the analysis result, and generates a corresponding analysis optimization report, and sends the analysis optimization report to the management personnel.

[0086] In one embodiment, first, based on the deviation data between the rainfall prediction and the actual data calculated previously, the root mean square error (RMSE) of the entire time period is further calculated. RMSE is a commonly used indicator to measure the difference between the predicted value and the actual value, which can reflect the overall accuracy of the prediction. After the calculation is completed, the root mean square error data of the entire time period needs to be counted and visualized in a preset manner. For example, a line graph can be used to show the trend of the root mean square error over time, or a bar graph can be used to compare the root mean square error size of different time periods.

[0087] By observing and analyzing the visualized root mean square error data, its changing trend can be determined. For example, if RMSE continues to rise in a certain period of time, this may indicate that the accuracy of rainfall forecasts in that period of time is decreasing. Based on the changing trend of RMSE, the performance of the numerical rainfall forecast can be evaluated in different time periods, and the time periods with low rainfall forecast accuracy and possible reasons can be identified.

[0088] According to the analysis results, the numerical rainfall forecast is optimized, which may include adjusting the parameters of the forecast model, introducing new data sources, or improving data processing methods to improve the accuracy of the forecast. After the optimization is completed, a detailed analysis and optimization report will be generated. This report contains the calculation results of RMSE, trend analysis, optimization measures, and evaluation of the optimized forecast effect. Finally, the analysis and optimization report is sent to the management personnel. This report not only provides an objective evaluation of the accuracy of the rainfall forecast, but also provides managers with specific suggestions for improving and optimizing the rainfall forecast.

[0089] The above is an embodiment of the method proposed in this application. Based on the same inventive concept, the embodiment of this application also provides a numerical rainfall forecast analysis device, whose structure is as follows: Figure 2 shown.

[0090] Figure 2 The internal structure diagram of a numerical rainfall forecast analysis device provided in the embodiment of the present application is shown in FIG. Figure 2 As shown, the device includes:

[0091] at least one processor;

[0092] and, a memory communicatively coupled to the at least one processor;

[0093] The memory stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor to enable the at least one processor to:

[0094] Collect boundary data of the area of ​​interest through multiple data sources, and convert the data format of the boundary data into a specified format;

[0095] Clip the boundary data of the specified format to determine the data of interest within the region of interest and generate the corresponding data set;

[0096] Based on the interest data in the data set, the rainfall forecast data corresponding to the time period to be executed is determined, and the actual rainfall data corresponding to the time period to be executed is obtained, so as to determine the deviation data between the rainfall forecast data and the corresponding actual rainfall data for each longitude and latitude point;

[0097] Based on the deviation data, the root mean square difference data of the whole time period is generated, so as to realize the numerical rainfall forecast analysis based on the root mean square difference data.

[0098] The present application also provides a non-volatile computer storage medium storing computer executable instructions. When the computer executable instructions are executed, they can:

[0099] Collect boundary data of the area of ​​interest through multiple data sources, and convert the data format of the boundary data into a specified format;

[0100] Clip the boundary data of the specified format to determine the data of interest within the region of interest and generate the corresponding data set;

[0101] Based on the interest data in the data set, the rainfall forecast data corresponding to the time period to be executed is determined, and the actual rainfall data corresponding to the time period to be executed is obtained, so as to determine the deviation data between the rainfall forecast data and the corresponding actual rainfall data for each longitude and latitude point;

[0102] Based on the deviation data, the root mean square difference data of the whole time period is generated, so as to realize the numerical rainfall forecast analysis based on the root mean square difference data.

[0103] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0104] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0105] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0106] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0107] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0108] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0109] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0110] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0111] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0112] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0113] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0114] The above is only the embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A numerical rainfall forecast analysis method, characterized in that: The method comprises: Collect boundary data of the area of ​​interest through multiple data sources, and convert the data format of the boundary data into a specified format; Clipping the boundary data in a specified format to determine the data of interest within the region of interest, and generating a corresponding data set; Based on the interest data in the data set, determine the rainfall prediction data corresponding to the time period to be executed, and obtain the actual rainfall data corresponding to the time period to be executed, so as to determine the deviation data between the rainfall prediction data and the corresponding actual rainfall data for each longitude and latitude point; Based on the deviation data, generating root mean square difference data for the entire time period, so as to realize numerical rainfall forecast analysis based on the root mean square difference data; The collecting of boundary data of the region of interest through multiple data sources specifically includes: When a preset scheduled task is triggered, the original NetCDF data is obtained based on web crawler technology, API interface call or geographic information service provider, and the original NetCDF data is stored in a pre-deployed OSS component; Obtaining the original NetCDF data from the OSS component to decompress the original NetCDF data to generate a corresponding sparse matrix, and reading boundary data corresponding to the region of interest based on the sparse matrix; The converting the data format of the boundary data into a specified format specifically includes: Cleaning the boundary data to remove duplicate data, invalid data or erroneous data in the boundary data, and determining a data format corresponding to the cleaned boundary data; Convert the cleaned boundary data into GeoJson format and wkt format, and store the cleaned and format-converted boundary data into a database; The step of determining the deviation data between the rainfall prediction data and the corresponding actual rainfall data for each latitude and longitude point specifically includes: Based on the corresponding relationship between the rainfall prediction data and the actual rainfall data, the rainfall prediction data and the actual rainfall data of each latitude and longitude point at the same time point are determined, and the deviation data between the rainfall prediction data and the actual rainfall data matched in time and space are calculated; wherein the deviation data includes a difference index and a ratio index; Comparing the deviation data with a preset deviation threshold to determine a corresponding magnitude relationship, and determining whether the deviation data is abnormal based on the magnitude relationship; If so, the outliers in the deviation data are removed, and the deviation data after the outliers are removed is smoothed to obtain the target deviation data.

2. A numerical rainfall forecast analysis method according to claim 1, characterized in that: The step of clipping the boundary data in the specified format to determine the interest data in the region of interest and generating a corresponding data set specifically includes: Based on the boundary data in the specified format, a cropping operation is performed on the region of interest to extract the data of interest in the region of interest; The intercepted interest data are sorted to generate a data set corresponding to the area of ​​interest; wherein the data set includes rainfall-related data corresponding to each latitude and longitude point in the area of ​​interest; The rainfall-related data are verified, and according to the verification result, the rainfall-related data are corrected to obtain a standard data set.

3. A numerical rainfall forecast analysis method according to claim 1, characterized in that: The step of determining the rainfall forecast data corresponding to the time period to be executed based on the interest data in the data set, and obtaining the actual rainfall data corresponding to the time period to be executed, specifically includes: Receiving a time period selection instruction input by a user through a preset visual interface, and determining a time period to be executed corresponding to the time period selection instruction; wherein the length of the time period to be executed does not exceed two hours; Based on the interest data in the standard data set, obtain the rainfall forecast data within the time period to be executed from the corresponding data source, and obtain the actual rainfall data corresponding to the time period to be executed from the actual observation site or the public rainfall data service; The rainfall prediction data and the actual rainfall data are temporally and spatially matched to determine the corresponding relationship between the rainfall prediction data and the actual rainfall data.

4. A numerical rainfall forecast analysis method according to claim 2, characterized in that: The performing spatiotemporal matching of the rainfall prediction data with the actual rainfall data to determine the corresponding relationship between the rainfall prediction data and the actual rainfall data specifically includes: The rainfall prediction data and the actual rainfall data are divided according to the same time granularity to determine the rainfall prediction data and the actual rainfall data corresponding to each time point; The rainfall prediction data and the actual rainfall data that are matched in time are spatially interpolated to determine the rainfall prediction data and the actual rainfall data corresponding to each spatial position, and the corresponding relationship between the rainfall prediction data and the actual rainfall data is determined in the case of time matching and space matching.

5. A numerical rainfall forecast analysis method according to claim 1, characterized in that: The step of generating root mean square error data for the entire time period based on the deviation data, and realizing numerical rainfall forecast analysis based on the root mean square error data, specifically includes: Based on the deviation data, the root mean square difference data of the entire time period is calculated, and the root mean square difference data of the entire time period is counted, so as to visualize the root mean square difference data in a preset manner; wherein the preset manner includes a line chart and a bar chart; Determining a change trend of the root mean square difference data, so as to analyze the numerical rainfall forecast based on the change trend; According to the analysis results, the numerical rainfall forecast is optimized, and a corresponding analysis optimization report is generated, so as to send the analysis optimization report to the management personnel.

6. A numerical rainfall forecast analysis device, characterized in that: The device comprises: at least one processor; and, a memory communicatively coupled to the at least one processor; The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a numerical rainfall forecast analysis method as described in any one of claims 1-5.

7. A non-volatile computer storage medium storing computer executable instructions, characterized in that: When the computer executable instructions are executed, a numerical rainfall forecast analysis method as described in any one of claims 1 to 5 is implemented.

Images